Comparison of two-dimensional manual and three-dimensional SurgiCase and BLUEPRINT planning software measurements of glenoid version, inclination, and humeral subluxation

Abstract

IntroductionVirtual planning for shoulder arthroplasty using preoperative computed tomography (CT) has been gaining popularity, and it is imperative for surgeons to recognize any differences in measurements that may exist amongst software platforms. The purpose of this study is to compare measurements of glenoid version, inclination, and humeral head subluxation between a manual approach and two varying automated software platforms using either a best-fit sphere technique (Wright-Medical BLUEPRINT) or an anatomic landmarks technique (Materalise SurgiCase). MethodsA case control study of 289 CT images from patients preoperatively planned for a total shoulder arthroplasty or reverse shoulder arthroplasty using SurgiCase (v3.0.110.5) were also successfully analyzed by BLUEPRINT (v2.1.6). Glenoid version, inclination, and subluxation were measured manually in a blind fashion by two separate investigators using axial and coronal images oriented to the scapular plane; interobserver and intraobserver reliabilities were measured using intraclass correlation coefficients (ICCs). Concordance correlation coefficients (CCCs), mean differences, and clinically relevant agreement in measurements between the software platforms and with the manual technique were compared. The impact of greater glenoid retroversion on the differences in measurements between the software platforms was further studied by correlation analysis. ResultsThe mean differences between SurgiCase and BLUEPRINT were + 0.5° for glenoid inclination (P = .064; CCC = 0.84), -0.9° for glenoid version (P < .001; CCC = 0.92), and -1.4% for humeral subluxation (P = .002; CCC = 0.88). Agreement within 5 units was 78.9% for inclination, 89.3% for version, and 64.1% for subluxation. Glenoid retroversion had no relation with the degree of variation in measured inclination (P = .59) or version (P = .56). There were significant differences between manual and 3D software measurements for glenoid inclination, version, and subluxation (P < .001). Both software measurements were more inferiorly inclined (average difference, SurgiCase -3.2° and BLUEPRINT -3.9°), more retroverted (average difference, SurgiCase -4.0° and BLUEPRINT -3.2°), and more posteriorly subluxated (average difference, SurgiCase + 3.4% and BLUEPRINT + 4.8%). ConclusionThe SurgiCase and BLUEPRINT preoperative planning software yield clinically similar measurements for glenoid version, inclination, and subluxation. The degree of glenoid retroversion does not impact the variability of inclination or version between the landmark and best-fit sphere software techniques. Compared to the 2D manual technique, both 3D software programs reported greater inferior inclination, retroversion, and posterior subluxation. Level of evidenceLevel III; Retrospective Diagnostic Study